Global Data Science Platform : Democratizing Data Sciences at Rakuten

14:20 - 14:40

Rakuten services generate a lot of data.  Rakuten Ichiba super-sales have generated logging data as high as 3.8 billion records per day, 120,000 queries per second. Internally, we have developed frameworks and tools to collect that data and make it available to a variety of internal capabilities that benefit our users and customers.  In this talk we will share will share our data collection infrastructure and talk about applications that make use of the data we have available.
Based on the user attributes and actions on the Rakuten eco-system, we have built a system that allows displaying users the benefits they will get upon their interactions with other diverse services . We will cover how we can retrieve data as well as how we can process data in real-time to be accurate, especially focusing on the functional and technical challenges this use case presents.